IBF-SLM Application Development

A prototype application for impact-based forecasting data processing, reasoning-ready corpus development, and model evaluation.

Stack

• FastAPI — async web framework
• SQLAlchemy 2 + asyncpg — async ORM with PostgreSQL
• Alembic — database migrations
• bcrypt + python-jose — password hashing and JWT tokens
• Jinja2 — server-rendered HTML templates
• Docker Compose — local PostgreSQL

Project structure

ibf_app/
├── app/
│   ├── core/
│   │   ├── config.py        # Settings from .env
│   │   ├── database.py      # Async engine + session
│   │   └── security.py      # bcrypt hashing, JWT helpers
│   ├── models/
│   │   └── user.py          # User model
│   ├── routers/
│   │   ├── auth.py          # /register  /login  /logout
│   │   └── dashboard.py     # /dashboard (protected)
│   ├── templates/           # Jinja2 HTML templates
│   └── main.py              # App entry point
├── alembic/                 # Migration scripts
├── docker-compose.yml
├── requirements.txt
└── .env.example

Getting started

1. Start PostgreSQL

docker-compose up -d

2. Configure environment

cp .env.example .env

Edit .env and set a strong SECRET_KEY:

python -c "import secrets; print(secrets.token_hex(32))"

3. Install dependencies

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt

4. Run

uvicorn app.main:app --reload

Tables are created automatically on first startup. No migration step needed for a fresh install.

Routes

Route	Description
GET /	Redirects to /login
GET /register	Registration page
POST /register	Create account
GET /login	Login page
POST /login	Authenticate, set session cookie
GET /dashboard	Protected dashboard
GET /logout	Clear session, redirect to login

Auth flow

Authentication uses httponly JWT cookies — no JavaScript required. On login the server sets an access_token cookie; the browser sends it automatically on every subsequent request. Visiting /dashboard without a valid token redirects to /login.

Database migrations

Alembic is wired up for production use. After changing a model:

alembic revision --autogenerate -m "describe your change"
alembic upgrade head